Building a robot is hard work. It's expensive. It requires expertise in a number of disciplines, and a willingness to make certain trade-offs for practical reasons. But many of those obstacles melt away when you can 3-D print a robot with liquids and solids simultaneously.
Researchers at MIT's Computer Science and Artificial Intelligence Lab have done just that. Their "printable hydraulics" method uses an inkjet printer to build up layer upon layer material less than half the width of a human hair one drop at a time. After each layer, a high-intensity UV light solidifies the material that is to be solid, while leaving the liquids liquid. In other words, the process can print a complex shape that contains liquid.
It's probably easiest to consider a practical example. Fortunately, the research team, led by lab director Daniela Rus and postdoctoral associate Robert MacCurdy, has one. The team printed a hexapod robot that weighs just 24 ounces and is less than 6 inches long. A motor spins a crankshaft, pumping fluid to the robot's legs. Several 3-D printed bellows use the fluid pressure of that pumping as a mechanical force, allowing the robot to move.
The result is a little plastic shuffling robot, sure. The important thing to remember, though, is that it's a little plastic shuffling robot that took just 22 hours to print, with relatively small material cost. All the team needed to do was slap a motor and battery on it, and it was ready to go. That's got exciting implications.
"The ability of printing solids and liquids at the same time will enable us to create a whole different class of active mechanisms," says Rus. "The idea of reducing or simplifying the amount of manual assembly that's required to create robots is really critical for getting broader adoption of robots, and making them more accessible."
In the long view, tech like this is an important step toward robots that walk out of a 3-D printer. For now, though, it's an important development for a robotics industry where progress often is slowed or stymied by the intricate nature of 3-D printing.
"Right now it takes years to make any kind of robot. You need to be an expert in mechanics, electronics, computation, software; you need a lot of expertise," says Rus. "With this tool you can think of your robot at a higher level, and can print the whole body without manually having to assemble it."
That also means faster prototyping and faster iterations. Think your robot might benefit from six legs instead of just two? It's much easier—and cheaper—to 3-D print those experiments than to fabricate them. In fact, using a 3-D printer could help reinvent the fabrication process altogether.
"If you look at a traditional robot, something that's designed and built in a conventional way, the designer of that robot has embedded all kinds of choices about the eventual fabrication approach that will be used," says MacCurdy. "We think that by coming up with a method that decouples the cost of fabricating from the design, we can enable new applications that we haven't even dreamed of."
Better still, there are many applications that don't require much dreaming at all. Being able to create a robot like this potentially allows for a new class of "disposable" machines that could enter hazardous areas, like nuclear sites, that conventional electronics may not be able to withstand.
Eventually, Rus says, the hope is that a process like this enables robots that can do anything from measuring radon to playing chess. There are plenty of challenges to overcome yet, mostly around how to incorporate the motor or muscle necessary to create propulsion. They're easy enough to add afterward, but the hope is that someday it will be part of the printing process. If and when that does happen, creating a robot won't be that much more difficult than creating doll. The end result will just be a whole lot more useful.
